Pension and Health Care Reforms in the Last Quarter Century. Carmelo Mesa-Lago

Transcription

1 S O C I A L S E C U R I T Y I N L AT I N A M E R I C A Pension and Health Care Reforms in the Last Quarter Century Carmelo Mesa-Lago University of Pittsburgh CRUCIAL NEEDS, WEAK INCENTIVES: SOCIAL SECTOR REFORM, DEMOCRATIZATION, AND GLOBALIZATION IN LATIN AMERICA. Edited by Robert R. Kaufman and Joan M. Nelson. (Washington, D.C. and Baltimore: Woodrow Wilson Press and Johns Hopkins University Press, Pp $55.00 cloth, $24.95 paper). EL SISTEMA DE PENSIONES EN CHILE EN EL CONTEXTO MUN- DIAL Y DE AMÉRICA LATINA: EVALUACIÓN Y DESAFÍOS. Edited by Oficina Internacional del Trabajo (OIT). (Santiago: Ponencias del Seminario Internacional, Pp. 174). HEALTH SECTOR REFORM IN BOLIVIA: A DECENTRALIZATION CASE STUDY. By World Bank. (Washington, D.C.: A World Bank Country Study, Pp. 94. $15.00 paper). INNOVACIONES EN EL SISTEMA DE SALUD EN AMÉRICA CENTRAL: LECCIONES E IMPACTOS DE NUEVOS ENFOQUES. Edited by Gerard M. La Forgia. (Washington, D.C.: Documento de Trabajo del Banco Mundial, No. 60, Pp $30.00 paper). KEEPING THE PROMISE OF SOCIAL SECURITY IN LATIN AMERICA. By Indermit Gill, Truman Packard, and Juan Yermo. (Stanford: Stanford University Press and World Bank, Pp $75.00 cloth, $39.95 paper). LEARNING FROM FOREIGN MODELS IN LATIN AMERICAN POLICY REFORM. Edited by Kurt Weyland. (Washington, D.C. and Baltimore: Woodrow Wilson Press and Johns Hopkins University Press, Pp $60.00 cloth, $22.95 paper). PROTECCIÓN SOCIAL Y MERCADO LABORAL. Edited by Fabio M. Bertranou. (Santiago: Oficina Internacional del Trabajo, Pp. 197). social security in latin america During the last quarter century, the most significant social policy transformation in Latin America has been pension and health care reforms. Total Latin American Research Review, Vol. 42, No. 2, June by the University of Texas Press, P.O. Box 7819, Austin, TX

2 182 Latin American Research Review or partial pension privatization has spread to twelve countries in the region, influenced similar changes in Central and Eastern Europe, and become a point of reference in the debate on reform in some Western European countries and the United States. Health care reforms have been implanted in all countries but with less impact abroad. The seven books reviewed in this essay address these issues; two of them also treat unemployment and education, reforms of lesser importance that will not be treated here due to space limitations. Five of the books are edited collections. In total there are more than sixty authors and thirteen countries involved, as well as different approaches to the reforms, therefore, it is impossible in this essay to do justice to these books, let alone touch on all contributions. The books deal with the following topics: the influence of foreign models in general, and specifically in the pension reforms of Argentina and Brazil, as well as the health reforms in Colombia and Mexico (Weyland); politics of health care reform in Argentina, Brazil, Colombia, Costa Rica, Mexico and Peru (Kaufman and Nelson); relationship between the labor market and employment with social security coverage, focusing on Argentina, Chile and Uruguay (Bertranou); evaluation by three World Bank experts of the results of structural pension reforms in ten Latin American countries in the last decade (Gill, Packard, and Yermo); Chile s pension reform within the Latin American context (OIT); health care innovations involving the private sector and their effects in five Central American countries (La Forgia); and health care reform in Bolivia (World Bank). pension reforms The 1994 World Bank report, Averting the Old Age Crisis: Policies to Protect the Old and Promote Growth, eventually became the world paradigm for structural pension reforms that totally or partially privatized public systems. The Chilean reform of 1981 preceded said report and together with the international financial institutions (IFIs: the World Bank, International Monetary Fund, and Inter-American Development Bank ) influenced similar structural reforms in nine other Latin American countries in : Argentina, Bolivia, Colombia, Costa Rica, Dominican Republic, El Salvador, Mexico, Peru, and Uruguay. Ecuadorian and Nicaraguan structural reform laws had not been implemented by the end of 2005 due to a constitutional court appeal and unsustainable fiscal costs, respectively. There are three general models of structural reforms: 1) substitutive (in Bolivia, Chile, Dominican Republic, El Salvador, Mexico, and Nicaragua), where the public system is closed and replaced by a private system; 2) parallel (in Colombia and Peru), where the public system continues, albeit reformed and a new private system is established and competes with the public one; and 3) mixed (in Argentina, Costa Rica, Ecuador, and Uruguay), where the public systems

3 review essays 183 continue as the first pillar of an integrated scheme while a private system becomes the second pillar and pays supplementary pensions. The other eight countries of the region (Brazil, Cuba, Guatemala, Haiti, Honduras, Panama, Paraguay, and Venezuela) retain public pension systems. A private system is characterized by defined contributions (supposedly unchangeable through time), undefined benefits (the pension is uncertain and will depend on factors related to the insured and the economy), fully funded financing based on individual accounts, and private administration. Conversely, a public system is characterized by undefined contributions (they tend to increase in the long run due to ageing of the population and maturity of the pension scheme), defined benefits (specified by law), pay-as-you-go PAGO financing (although four Latin American public systems have reserves and operate on partial collective funding), and public administration (Mesa-Lago 2004, 2006a). Another major difference is that in the capitalization [private] system both the losses from economic downturn and the gains from development are individualized and appropriated [while] in PAGO systems, the benefits and costs are socialized and subject to redistributive rules defined by the government (Vinicius Pinheiro in Weyland, 115). Weyland convincingly argues that Chile became a model of pension reform because it surged at a time of mounting criticism on the problems in public systems, which seemed impossible to solve within the prevailing paradigm. Chilean reformers substituted a bold new paradigm (privatization), inserted into the framework of neoliberal policies, and promised both to solve said problems and achieve broader long-term goals (converting the fiscal deficit into national savings that would promote capital markets, economic growth, and employment), thus mobilizing additional supporters. He also argues, less successfully, that the high status of Chile as a nation contributed to turn its policy into a regional model, while cultural and socioeconomic similarities enhanced the probability of its emulation. But Chile was an international political outcast while under the rule of Augusto Pinochet dictatorship. Weyland adds that the Chilean model became more attractive when the country returned to democracy, yet maintained the pension scheme (11). Actually, the first country to follow Chile s substitutive model was Bolivia in 1997, despite significant socioeconomic differences with Chile. Weyland rightfully notes that the enormous transitional costs of Chile s full privatization induced other countries to scale down the bold model and adopt parallel or mixed systems (in ), easier to finance and get political approval in democratic countries. His rationale, however, does not explain the almost full imitation of the Chilean model in other democracies such as El Salvador, Mexico, and the Dominican Republic. Diffusion was helped by third parties, such as IFIs, that may push a model that conforms to their normative or ideological orientation, but

4 184 Latin American Research Review does not fit the specific needs of the recipient country, and yet what worked in one country may not work in another; it may even make things worse (23). The Chilean model was oversold, imitated before it started to pay benefits to a substantial number of people, and hence was a matter of projection or speculation not fact (Weyland, 10, 273). The assumptions and actual effects of pension structural reforms are analyzed in the light of statistics from the ten Latin American countries with reforms, the report of the World Bank, and other four books reviewed herein, as well as additional pertinent works. The World Bank report ratifies most key assumptions and recommendations of the previous report but questions some of them, and identifies and analyzes fundamental problems confronted by the reforms after one decade in operation (Gill, Packard, and Yermo, hereafter GPY). Have the reforms been successful? The report s answer is that it depends: If the new structures are viewed as a final design, the [reforms] may well be assessed harshly, because scores of people are left uncovered... there are still some adverse equity effects, and the cost and risk management features are somewhat deficient. On the other hand, if the current structures are viewed as a transitory stage, [the] reforms should be viewed as successful, because the movements have been in the right direction. Despite the problems found, the report argues that it would be a mistake to go back to the unsustainable structures that existed [before], however, it would also be a mistake to think of the [current] stage as the final structure (GPY, 13 14). The reforms were expected to extend coverage of the labor force through an increase in incentives for affiliation. According to the World Bank, despite an improvement of such incentives in most countries, coverage, after rising modestly, has stalled at about half of the labor force in two countries (Chile and Mexico), but in most countries coverage is much lower; stagnant coverage is indicative of skepticism of the new system and may even signal a rejection by many workers (GPY). My own estimates indicate that coverage has declined rather than stagnated: comparing the year before the reform and 2004, coverage decreased in all ten countries, and total average coverage fell from 38 percent to 26 percent of the labor force. The World Bank assertion that coverage rose modestly in Chile after the reform fails to note that coverage was higher prior to the military dictatorship and declined to a trough in 1980, the year before the reform. Combining private and public schemes, such coverage decreased from 62 percent in 1975 to 57 percent in 1991, and rose to 61 percent in 2003, still below the pre-reform level (Arenas de Mesa in OIT 2004). The reformers predicted that incentives such as the ownership of the individual account and the strong linkage between contribution and the pension level would promote punctual payment of contributions. The

5 review essays 185 World Bank acknowledges, however, that despite the time the reforms have been in operation, the expected improvement in incentives has not been rigorously tested. Data show that the percentage of affiliates who were active contributors steadily decreased in all countries in , with few exceptions, and the average number of affiliates that contributed fell from 58 percent to 42 percent (Mesa-Lago 2006a). A major problem confronted by all social security systems, regardless of their private or public nature, is the profound transformation in the labor market that occurred in the last twenty-five years, an issue analyzed in the book on social protection and labor market (Bertranou). Social security was originally designed for salaried workers in the formal sector and with stable jobs, and financed by contributions from employers and employees. But since the crisis of the 1980s in Latin America, the formal sector has shrunk and the informal sector has expanded. Informal workers, not covered by social security, include the self-employed, unpaid family workers, domestic servants, and employees in microenterprises. In addition, neoliberalism and globalization have promoted labor flexibilization, a proliferation of jobs that also lack social protection, such as workers without a contract or under subcontracting, part-time, and temporary work. In , coverage of workers by social security experienced an overall decline, but still those in the formal sector had four times more coverage that those in the informal sector (Bertranou 2004). Social security has not adapted to such drastic changes, and the most important challenge it faces is to stop the decline in protection of the formal sector and expand coverage to the growing informal sector. 1 There is need for new approaches to incorporate the self-employed and other workers in the informal sector through formalization processes, legal obligation to affiliate, fiscal incentives and/or solidarity funds, facilities for registration and payment of contributions. But compulsory coverage cannot solve the problem by itself: most self-employed have unstable jobs and low income, and they are required to contribute both for themselves and for the employer that they lack, hence the need for fiscal subsidies. Only six countries in Latin America provide social assistance pensions for the poor, and they are the pioneers with the highest coverage and usually the lowest poverty incidence: Argentina, Brazil, Chile, Costa Rica, Cuba, and Uruguay. But these pensions are not provided to all the needy, their level is often insufficient to cover basic needs, they are usually submitted to quotas and waiting lists, the amount of the pension depends on available fiscal resources, the qualifying age is higher than for contributory pensions, and targeting the poor needs improvement. Despite 1. In his speech at a seminar held in Santiago de Chile on November 11, 2005, Minister of Labor Yerko Ljubetic stated, It is not the real world that should be tailored to the pension system [as administrators of such system pretend] but the system that must be adapted to reality.

6 186 Latin American Research Review these limitations, social assistance pensions have significantly reduced poverty in those countries and only cost from 0.2 percent to 1 percent of GDP (Bertranou et al 2002). Costa Rica s reform law of 2001 mandated a social assistance pension to all who are 65 and older and lack resources, but it has not been implemented yet (Martínez and Mesa-Lago 2003). The remaining fourteen countries lack social assistance pensions, although they endure the highest poverty incidence. Bolivia grants a small flat annual sum to workers who were 21 years and older in 1995, regardless of income, leaving later generations without poverty protection (GPY). Laws in Colombia, Dominican Republic, and Ecuador stipulate social assistance pensions but had not been enforced in The World Bank recommends that priority be given to poverty prevention (first pillar), instead of to mandatory savings (second pillar), as was done in the region in the last decade. There is an increasing consensus on this issue, regardless of the divergent ideological position of the experts (see OIT 2004). As a result of declining coverage of the active labor force, protection of the elderly is also dwindling, a trend that will worsen in the future unless coverage of both contributory and assistance pensions is extended. According to household surveys taken in , three countries had percent of the aged protected by contributory pensions, while in other six countries protection was only 9 20 percent, and in several countries the share of the elderly receiving pensions was falling (GPY). Structural reformers claimed that private systems would improve efficiency and reduce administrative costs by breaking the monopoly of public systems, introducing competition, and granting freedom to the insured to select and change administrators based on information on their commissions, rates of return, etc. But competition is afflicted by important flaws: heavy regulations and tight restrictions on changing administrators have created a captive clientele for each pension fund administrator and institutionalized what was de facto already an oligopoly, and evidence from Latin America clearly shows that the current pension fund industries are anything but good examples of competition (GPY, 233, 238). Competition largely depends on the size of the insured market: the bigger it is, the more administrators operate and vice versa. Thus, in 2004 Mexico had 32 million affiliates and 12 administrators, Chile 7 million and 6, but Bolivia and El Salvador about 1 million and 2 (in Bolivia the government assigned the insured among the two administrators based on their place of residency and banned changes among the two until 2003). In the long run the number of administrators diminishes: in Argentina from 25 to 12 and in Chile from 21 to 6. Small countries will face significant obstacles to secure a sufficient number of administrators and competition. Concentration is significant and growing: in 2004 it was 100 percent in the biggest three administrators in two countries and 71 percent to 86 percent in another four countries; concentration in Chile s biggest three rose from

7 review essays percent to 80 percent in (Mesa-Lago 2005). The assumption that the insured have proper information and skills to make an educated selection among the best administrators is denied by surveys that show astonishing lack of knowledge. Contrary to the argument that privatization would reduce managerial costs, the shift from one public to multiple private administrators has provoked losses in economies of scale, substantial expenditures on publicity, costly commissions paid to salespersons that have generated high turnover among affiliates, and lofty profit margins (Pinheiro in Weyland). Furthermore, it has been shown that competition does not work properly in most private systems. The World Bank argues that Latin American private pension systems have been generally successful at reducing costs, although with three important caveats: commissions are still unacceptably high for a large percentage of the population, only a small portion in the cut of operating expenses is transferred to affiliates as lower commissions, and such diversions from worker contributions that are mandated by governments should be cause for concern (GPY, 8, 233). There is strong support now in Argentina, Chile, and Uruguay to control administrative costs, and the World Bank sets as a priority the need for private systems to cut their costs. The regional rate of capital accumulation in the pension funds doubled from 7.1 percent to 13.5 percent of GDP in (GPY). The accumulation and rate, however, vary greatly among countries due to the size of their economies, the number of insured, the wage level, rates of return and the time the private system has been in operation. The World Bank admits that the only way to capital accumulation is not through the heavy reliance on a mandated second pillar (as in the ten countries with structural reform); countries such as Brazil that have reasonably well-developed capital markets may well choose to change the parameters of their public PAYG pension system rather than switch to a mandatory funded program (GPY, 13, 277). Due to the high fiscal costs of the transition, many governments have set ceilings on investment instruments, forced pension funds to invest a minimum percentage in public-debt bonds and prohibited or restricted foreign securities. Therefore, financing to the private sector through bonds and equities is still relative low. Pension funds are the dominant investor in capital markets and are increasingly concentrated into fewer hands. The precarious fiscal positions of the governments in the region have through high interest rates on government debt instruments resulted in high rates of return of the portfolio, which raises three concerns: how long these high returns can be maintained, the risk of default, and falling returns (GPY). 2 In 2004, four out of ten countries had from Annual real average gross rates of return since the beginning of the system until the end of 2004 were: 13 percent in Uruguay; 10 percent in Argentina, Bolivia, Chile and El

8 188 Latin American Research Review percent to 86 percent of the total pension fund invested in government debt and 55 percent invested in government debt in two other countries. Investment in stocks averaged only 8.4 percent and was only significant in four countries; the alternative to invest in foreign instruments was prohibited in several countries; the share averaged 5.6 percent and was significant in only four countries (AIOS 2005; Mesa-Lago 2006a). Small countries without a capital market or an incipient one confront a severe barrier to diversify their portfolios and a high risk of heavy dependence on public debt instruments. Fiscal costs of the reform are difficult to measure, project, and compare among countries because of diverse components and methodologies. The World Bank estimates in 2001 are higher than the corresponding domestic estimates in Argentina, Bolivia, and Colombia, and higher than the projections made at the start of the reform. Despite the projected cut in the fiscal debt generated by the reform, fiscal sustainability is far from assured, and empirical evidence shows that pension reforms can produce severe cash-flow problems in excess of initially projected transition costs (GPY). Countries considering a structural reform should undertake projections of fiscal costs in a realistic and cautious manner, with the aid of international organizations; publish such projections for public scrutiny; and determine how to pay for the fiscal costs during the transition. Structural pension reforms have worsened gender inequality for a variety of reasons. First, most private systems have increased the number of years of contribution required to grant a minimum pension, thus making it more difficult for women to earn that benefit. Second, the pension is based on contributions during the whole working life, instead of only the last years before retirement, as in most public systems in the region, harming women whose density of contributions is low. Third, the annuity is calculated with mortality tables differentiated by gender and the sum accumulated in the individual account is divided by the average life expectancy. Hence, women s pensions are lower than men s, who have a shorter life expectancy (Bertranou and Arenas de Mesa 2003). The World Bank report assesses the gender impact of the reform in eight countries based on the difference between internal rates of return of women and men, with mixed results, but acknowledges that in all countries, despite the reforms, women earn lower returns than men. Also, the report admits that due to higher life expectancy and sex-specific mortality tables, annuities received by women are lower than men, even if they retire at the same age (GPY). Among policies to improve gender equality are using unisex mortality tables, mandating Salvador; 6.7 percent to 7.7 percent in Costa Rica, Mexico and Peru, and -9 percent in the Dominican Republic (AIOS 2005). After the subtraction of administrative costs, the net rates of return were smaller. In Chile the average rate of return declined from 10.4 percent in to 5.1 percent in (Arenas de Mesa in OIT 2004).

9 review essays 189 that an insured married male who retires does so with joint annuities that would cover his female spouse, and giving women credit for raising their children as was done in Chile prior to the reform. One alleged advantage of private systems is that they have a defined contribution, i.e., the percentage paid upon the salary will not change in the future, in contrast with the public system s tendency towards increasing contributions, which results from population ageing and maturity of the scheme. But demographic factors will affect private systems also: the rising life expectancy will eventually force either an increase in the contribution or a cut in the pension level (as the amount accumulated in the individual account will have to be stretched over a longer period of retirement) or an increase in the age of retirement or a combination of those measures (Nicholas Barr in OIT 2004). Contrary to the widespread assumption that the insured ownership of the individual account combined with the private administration of the pension funds would impede the traditional state and political interference in public pension systems in the region, the World Bank states, The ability of the multi-pillar model to isolate the pension system from abuse by governments may have been oversold by reformers... the degree of protection against policy risk offered by privatizing... pensions can be exaggerated... the crisis in Argentina illustrates how any government organized retirement security system... can fall prey to politicians [but] similar threats to the viability of funded pension schemes can emerge in other countries of the region (GPY, 5, 133). The above discussion has been restricted to private systems with structural reforms. What is the situation in the eight countries that still have public systems? Nonstructural or parametric reforms geared to financially strengthen public systems (by raising ages of retirement, tightening benefits, increasing contributions or a combination of those measures) have been recently implemented or approved in three public systems. Brazil reformed the scheme for private workers in , tightening the rules for eligibility of benefits and inserting a notional defined contribution financial regime, 3 and also reformed the schemes for public employees in 2004, gradually eliminating privileged benefits, thus improving equity and reducing fiscal costs (Helmut Schwarzer in OIT 2004; Vinicius Pinheiro in Weyland). Venezuela passed a structural reform that was abolished by the current government, which in turn approved parametric reforms in , some still not in operation. Costa Rica s first (public) pillar was reformed in 2005, after eight years of postponement. The pension for those who retire before the normal age of 65 (in 3. This scheme is usually unfunded, contributions are credited to individual accounts but there is no capital accumulation and the state is responsible for paying the pension, the level of which is often set based on life expectancy, ageing of the population and decline in the active contributor/ pensioner ratio.

10 190 Latin American Research Review practice 57 for women and 62 for men) is proportionally reduced according to the age of retirement, hence substantially expanding the period of actuarial equilibrium. In Panama a parametric reform that was based on an International Labour Organization study and had been approved by workers, employers, and the government back in 1998, was halted by the government in The new government that took power in 2005 submitted a proposal to congress raising the contribution rate and adjusting benefits, but public opposition halted the necessary reform (ISSA 2005). Cuba s pension system confronts severe financial problems: it has the oldest population in the region after Uruguay, ages of retirement are among the lowest (55 for women and 60 for men), life expectancy of pensioners is the longest, the employers contribution is insufficient to finance benefits, less than 15 percent of workers contribute, and the state has to cover a growing deficit (from 1.3 percent to 2.3 percent of GDP in ), and pensions fell sharply in real terms. Pensions increased somewhat in 2005, but the system is unsustainable without a parametric reform (Mesa-Lago in Witte 2003). Guatemala, Honduras, and Paraguay have considered both structural and parametric reforms but have not taken action. health care reforms Latin American health indicators have improved substantially since at least the 1960s, but such improvements have been uneven in the region and insufficient to reduce pervasive social inequalities and provide effective access to the poor. At the start of the 1990s, health care systems suffered deterioration and fiscal constraints from the previous decade crisis. They were highly centralized and segmented among and within three sectors: public, social insurance and private. They also responded largely to organized labor s needs and were afflicted by low access, equity and quality, inefficiency, and inequitable allocation of resources among population groups, regions, and levels of care. The reforms of the 1990s attempted to cope with those problems and improve equity, efficiency, and quality, although efforts and results varied widely (see the useful comparative table of Nelson in Kaufman and Nelson, pp ). By the end of 2005, virtually all twenty countries had introduced some type of health care reform, in contrast with only ten structural pension reforms, but the former have been less profound and with significantly more differences regarding scope, depth, progress, and features. It is therefore more difficult to identify general types of health care reforms than pension reforms, where a distinction has been made between structural and parametric. Because of the substitution of the public by a private pension system under structural reform, from 87 percent to 100 percent of the insured in eight of the ten countries are under private provision. In health reforms, such distinction is difficult to establish because the

11 review essays 191 private sector covers only from 13 percent to 25 percent of the population in the countries with the most advanced degree of privatization. Due to the diversity of health care reforms, ten different models have been identified (compared with only three models in pension reforms), with divergent degrees of coverage, integration/coordination, decentralization and hospital autonomy, separation of functions, competition, freedom of choice, privatization, efficiency, shift in fiscal subsidies from supply to demand, financing, social participation (Mesa-Lago 2006b). The only global, large-scale, and fully implemented health care reform occurred in Chile. Similar reforms in Brazil and Colombia remain far from complete, and in most countries the reform has been piecemeal, targeted to a specific problem or aspect of the system, and incomplete, with the exception of Costa Rica (Nelson in Kaufman and Nelson). In , eight countries approved or were discussing legislation or modifications in their health care systems. The 1993 World Bank World Development Report: Investing in Health exercised considerable influence in health care reforms in the region, but with less impact than its pension counterpart, because it failed to turn into a clear, comprehensive, and integrated paradigm commanding widespread support to challenge the prevailing paradigm. Hence more latitude was left to the countries for designing their own reforms (Nelson in Weyland). Chile s global-radical reform (initiated in 1981 and completed in 2005) attracted wide attention but very few imitators (Kaufman and Nelson), and none of the books under review deals with it (see Bortzutzky 2003; Homedes and Ugalde 2005; Urriola 2005). This section focuses on the global-radical reforms of Brazil and Colombia, and piecemeal-moderate reforms in Argentina, Central America and Bolivia (analyzed in four of the books). Reforms in Mexico (discussed in two of the books) and Peru (analyzed in one book) are piecemeal and incomplete and moderate respectively (see the taxonomy in Kaufman and Nelson). Uruguay remains virtually without reform and is excluded by all the books. The emphasis is on the type of reform and its effects (positive and negative), rather than on foreign influences and the politics of the reforms (major topics in Kaufman and Nelson; Weyland). The Colombian reform that began in 1993, which is probably the most global and complex in the region, provides health insurance through two regimes: contributive for those in the formal sector capable of contributing (coverage was extended to their dependent relatives), and subsidized for those with insufficient resources, who didn t have insurance coverage before; the target to have all the population insured by 2001 was not met. Patricia Ramírez (in Kaufman and Nelson) and Juan Pablo Uribe (in Weyland) assess the reform outcomes, positive and negative (additional data and comments below are from Mesa-Lago 2006b). Health insurance coverage of the population increased from 21 percent in 1993 to 57 percent

12 192 Latin American Research Review in 1997 but decreased to 53 percent in ; while contributory coverage has shrunk, subsidized protection has expanded. Uribe estimates that 22 percent of the population the poorest lacks insurance, although legally they have access to public hospitals; and yet there is a residue of percent that cannot be explained solely by those insured in separate schemes (armed forces, teachers and petroleum workers), hence the percentage of uninsured must be higher. Equity has improved because subsidies to demand (rather than supply) have allowed targeting a basic package on the poor and increased their access, as well as that of the lowest-income quintiles, the rural population and the self-employed. Nevertheless, about 40 percent of the population (poor and low income) still does not receive subsidies and continues under the prereform system, while 30 percent of the beneficiaries of the subsidized regime are not poor and the 2001 goal of an equal benefit basic package under the two regimes did not materialize. A solidarity fund created with a percentage of salary in the contributory regime, national budget transfers, and petroleum taxes, allocates resources to the subsidized regime, but its revenue sources have dwindled, and delays and confusion have obstructed the transfers. Health expenditures rose from 6.8 percent to 9 percent of GDP in (but declined to 5.5 percent in due to the economic crisis); distribution of such expenditures is better because more is allocated to primary care over high-complexity hospitals. However, a significant part of the new resources intended for expanding coverage has gone to the bureaucracy, wage increases, and corruption. The system is besieged by high administrative costs associated with complex and poorly transparent financial flows, the lack of an integrated information system, weak supervision, and opportunism of some public and private providers. According to Uribe, the quality of services is better as measured by consumer satisfaction surveys although he does not provide data; and yet, the level of satisfaction was ranked a meager 6 in a range from 0 to 10 in , and 62 percent of health professionals considered that the general level of care had declined. Infant and maternal mortality have continued to decline, although a positive correlation cannot be established with expansion of insurance coverage. Vaccination coverage has dropped drastically, however, and some epidemics have reemerged. The Brazilian reform, implemented in the 1990s, unified a segmented public sector largely based on social insurance into a unified, integrated and free public health system (SUS) divided into three levels: federal (coordinator and main financier), states (regulator and partly financier) and municipalities (managers of the basic package and family health program). Nevertheless, significant heterogeneity persists. The private sector (regulated and supervised by the state) is the largest in the region and growing; it covers close to 25 percent of the population (versus 18 percent in Chile) but has 70 percent of all hospital beds, 58 percent of them subcontracted

13 review essays 193 with SUS. Coverage is apparently universal, although there are no accurate data on access. The poor and low-income population uses SUS (at least 7 percent of the poor lack access) and the middle- and high-income groups use private providers, but part of them resort to SUS for costly hospital services. Primary care expanded from 73 percent to virtually 100 percent of the population in , while the family health program jumped from 0.7 percent to 33 percent. Surveys indicate increasing satisfaction among users but more in primary care than in hospitals. Total health expenditure rose from 7.9 percent to 8.5 percent of GDP in , the public share was stagnant but the private share consistently expanded; there was a significant increase in the allocation to primary care. Despite this impressive progress, Brazil ranked in 2000 as the worst country in the region in terms of financial equity and, although it was fifth highest in health expenditure per capita, it ranked tenth in health status indicators (Mesa-Lago 2006b). Marta Arretche (in Kaufman and Nelson) explains part of that paradox. First, coverage by SUS decreases with income and education while the opposite is true with coverage by the private sector. Second, national average health indicators have improved significantly but they remain very uneven among regions: infant mortality in the wealthy Southeast is one-third that in the poor Northeast, whose poorest municipalities have rates similar to those in Africa while the richest municipalities in Rio Grande do Sul (in the Southeast) have levels similar to those in Costa Rica, and Chile life expectancy exhibits similar trends. Third, SUS spending on hospital care is still very concentrated in the Southeast, superseding its population share. Finally, SUS allocated 56 percent of total hospital spending to private and philanthropic institutions (most of the latter are actually profit seekers), which are the best equipped and are mainly located in the Southeast. The Argentinean reform of the 1990s was initially heralded as a radical break with the past and a regional model of successful reform. Peter Lloyd-Sherlock (in Kaufman and Nelson) argues that the reform has not been completed. Rather, it is a collection of separate components that lacks coherence; has increased previous fragmentation and complexity; and has failed to implement effective regulation, coordination, and supervision, thus reducing the prospects for a more unified system in the future. And while he notes some significant positive effects of the reform, he argues that the crisis has affected both the federal and provincial governments, and OS (Obras Sociales social insurance mainly managed by trade unions) and other providers are significantly indebted to the hospitals. The reform has failed to regulate the private sector, allowed the levy of voluntary user fees that had been previously eliminated, and has not solved the crisis of the OS for pensioners (PAMI). In 2002, a health sector emergency decree put in doubt the future of the reform.

14 194 Latin American Research Review La Forgia s compilation focuses on lesser-known, piecemeal but important health reforms ( innovations ) that attempt to tackle a particular problem in Central America (El Salvador is excluded). Despite progress achieved in the last two decades, most of these countries lag behind the rest of Latin America in their health care status (Costa Rica and Panama are exceptions). All but one deal with primary care provision, and all try to improve incentives through delegation to private providers or to a public-private mix (mostly NGOs and cooperatives rather than profit-seeking enterprises), as well as some degree of competition, but without significantly curtailing the dominant public sector. The final chapter compares features and effects of the five innovative programs with those of the traditional providers ministries of health and social insurance with the following results: positive for the innovations in Costa Rica and Honduras, mostly positive in Panama, positive but with deficiencies in Nicaragua, and mixed in Guatemala. At the end of the 1980s, Costa Rica s social insurance institute (CCSS) had achieved virtually universal coverage and significantly improved health standards, which probably were the highest in the region. The government introduced a reform in the early 1990s, justified by insufficient CCSS infrastructure, high and increasing costs, restrictions to hiring extra personnel, a decline in quality and user satisfaction, and growing demand for services. Arguing that CCSS facilities were inefficient in coping with demand, the CCSS began to contract services (mainly at the primary level) with private providers. And yet, the actuarial reserves of the CCSS health scheme might be excessive and part of them could be invested in the needed infrastructure and equipment to cope with the insufficient internal capacity, instead of delegating provision to the private sector (Martínez and Mesa-Lago 2003). The gradualist reform in the last fifteen years has been dramatically different than privatization in other countries: the CCSS guarantees the services and retains responsibility for their financing, regulation, and supervision; primary services of the ministry of health were integrated into those of the CCSS and transferred to decentralized EBAIS (Equipos Básicos de Atención Integral en Salud) at the local level, which operate in all municipalities. The CCSS contracts provision of part of such services mostly from cooperatives, and also from the national university and private clinics. 4 Contrary to the expected reduction in costs, four evaluations conducted in showed that cooperatives had higher costs than CCSS clinics. The book raises questions about these evaluations and undertakes a comparison of results between both types of providers (three each from cooperatives and CCSS clinics, similar in population 4. An analysis of the success of EBAIS and the slow progress of deconcentration and hospital autonomy is done by Mary Clark (in Kaufman and Nelson)

15 review essays 195 area, level of complexity, etc.) based on indicators of performance and their evolution in The results indicate that the cooperatives delivered more visits per capita of general medicine and fewer specialist visits, while the number of emergency visits and primary care patients were the same. They also showed that cooperatives performed fewer lab tests, at 30 percent lower average cost per capita than the CCSS clinics. Surveys among cooperative users showed that 62 percent were satisfied with their services and 68 percent said that waiting time was shorter than in the CCSS clinics (James Cercone et al in La Forgia). Panama s health system has a high degree of integration and nearly universal coverage, mostly through social insurance (CSS) and, to a lesser extent, by the ministry of health, with a tiny private sector. CSS provides integral benefits, and the ministry a basic package only in two regions. In 1998, the reform created in an area of the capital city an autonomous entity, CONSALUD, that receives annual budgetary allocations from the ministry and social insurance in exchange for providing integral services to users in the area by an integrated hospital, HISMA. HISMA in turn subcontracts all services medical, administrative, alimentary, and cleaning in triennial public bids among private providers. CONSALUD pays HISMA predetermined fees based on cost studies for each service, sets ceilings for the production of its services to control expenditures, and audits its bills. HISMA payments to providers depend on an evaluation of the quality of their services and users satisfaction measured by monthly surveys. In 2000, a comparative study on efficiency with 23 indicators to measure performance between HISMA and two hospitals with similar features (one each from the ministry of health and social insurance), HISMA got the highest marks in accreditation of quality and user satisfaction but the lowest in coverage of the population and bed utilization. It also received the highest in productivity of physicians and nurses in emergency care, but the worst in ambulatory consultation. HISMA also performed better in half of ten cost indicators but worse than one or both hospitals in the other half of the cost indicators. The study also detected two problems in the model: 1) a monopsony (CONSALUD is the only buyer of services) and a monopoly (HISMA is the only supplier of services to the users, albeit provided by competing enterprises), that could be negatively affecting the potential gains in efficiency of this model ; and 2) HISMA receives a uniform tariff for each hospital release, ambulatory surgery, and emergency consultation, which could incentivize the selection of the healthiest patients in each service. The extension of this model to other hospitals has been halted (Ricardo Bitrán et al in La Forgia: ). The health care reform that Bolivia began in 1994 is more global than those in Central America but less than those in Brazil and Colombia. The World Bank (2004) documents some improvement in several areas of Bolivia s accumulated health problems. According to the Bank,

16 196 Latin American Research Review coverage increased to percent in 2000, due to a jump in SUMI (Seguro Universal Materno-Infantil, a basic health package) protection from 30 percent to percent; however, 79 percent of the population lacks any type of insurance, and social insurance covers only 4 percent of the population in rural areas, where the indigenous groups are concentrated. The self-employed have voluntary coverage and must pay their contribution plus the employer s, a serious barrier to their coverage. In 2001, social insurance covered 5.8 percent of the poorest population quintile, versus 31 percent of the richest quintile. The Administrative Decentralization Law of 1994 divided public-sector functions into four levels, the ministry, regional departments, municipalities, and districts, a fragmentation that provoked a variety of problems. The ministry is supposed to monitor performance agreements signed with the regions and these in turn with municipalities, but there are neither precise criteria to determine the targets, nor a clear methodology to measure performance, nor regular evaluation of the agreements. Still, the Bank reports an improvement in fixing targets and priorities, as well as progress in some indicators in recent years. Despite a decade of decentralization, 80 percent of total public expenditures in 2000 were still by the central government, and only 20 percent were by the municipalities. The ministry continues to manage human resources, and inefficiencies continue: loss of personnel time in nonproductive activities, unjustified absences, overtraining, frequent strikes, a national turnover rate of percent annually, and public care concentrated in the mornings because 30 percent of physicians work for only three hours a day. The World Bank argues that equity has improved through SUMI expanded coverage, benefits, and financing in But significant geographical and cultural access barriers remain, and SUMI does not reach the poorest populations, isolated rural areas and most indigenous people. Total health expenditures averaged 4.5 percent of GDP in and rose to 5.2 percent in Still, Bolivia s health expenditure per capita was the lowest in the region after Haiti in The percentage of public expenditures allocated to salaries increased from 51 percent in to 81 percent in At the same time, however, the percentage allocated to investment fell from 15 percent to 6 percent, and that to medicine also shrank, thus transferring the burden to the households. Furthermore, increased financing has not significantly reduced inequality gaps among municipalities, and total expenditure is unequally distributed among the three health sectors: 44 percent goes to social insurances that cover percent of the population, percent to the private sector that covers 5. Most of the 13 social insurance programs cover middle and high-income groups (university professors, armed forces) and receive cross subsidies, e.g., 45 percent of birth deliveries of social insured women are performed at public institutions.

17 review essays percent (most of it is family expenditure), and only 22 percent to the public sector that should cover 65 percent (Mesa-Lago 2006b). According to the World Bank, public expenditure was 20 percent of social insurance expenditure per capita in conclusions Private pension systems face the following challenges: neglect of poverty prevention (first pillar) and excessive emphasis on compulsory capitalization (second pillar); decline in coverage of both the active labor force and the elderly, aggravated by the transformation of the labor market (a trend clearly visible since the 1980s and overlooked by reformers); failure of incentives for the insured to contribute and pay punctually, resulting in growing noncompliance; malfunction of competition in most countries, leading to high and sustained administrative costs; lack of adequate information, ignorance of key aspects of the system, and poor skills of the insured to select the best administrators; worsening of gender inequality, an issue also totally neglected by the reformers; and the dream that a private system was immune to political interference shattered by the Argentinean crisis. The deceptive early success of Chilean pension reform and its widespread diffusion, combined with the financial and technical support of IFIs, hastened several countries to imitate a radical full privatization reform that led to some unsatisfactory results. A better-informed, cautious, and gradualist approach, taking into account the needs and characteristics of each country, including its level of socioeconomic development, and with more accurate projections of the reforms fiscal effects could have helped reformers avoid many problems. Knowledge of some public systems that have performed better than private systems in key indicators could have led to the implementation of appropriate parametric reforms in some countries, and notional defined contribution systems (as in Italy, Sweden, Poland and Brazil). Nations like Bolivia, Dominican Republic, El Salvador, Nicaragua, and Peru have a larger informal sector than Argentina, Chile, Costa Rica, and Uruguay and therefore face the lowest coverage and highest barriers for extension. Countries with considerably smaller domestic capital markets than Chile, which are all of the countries mentioned above, except Peru, plus Costa Rica and Uruguay, have found it impossible to diversify the portfolio of pension funds. The problem was aggravated in some of these countries when governments banned investment in foreign instruments. Some reforms were explicitly undertaken in pursuit of increasing national savings despite the fact that there is no solid evidence of that effect in Chile after twenty-four years. The highest capital accumulation in pension funds in the region is in Brazil, which has a public system but voluntary supplementary pension schemes. The small insured market in several countries is a barrier to competition,

18 198 Latin American Research Review which is the foundation of the private system (Bolivia copied Chile s model but with a virtual monopoly). Chile has generated a significant fiscal surplus to tackle the heavy, prolonged costs of the transition, an advantage that Argentina and other countries lacked, thus generating pressures to invest most of the pension fund in public-debt instruments. Nicaragua had to postpone indefinitely the implementation of its reform due to expected high transitional costs, while other countries (Bolivia and Peru) sacrificed benefits of the insured during the transition to cut transitional costs. Initial projections of such costs were too optimistic, reality has amply superseded them, creating new fiscal pressures. Ecuador and Nicaragua have been unable to implement their reforms, and the Dominican Republic has postponed the realization of its second and third stages. Argentina, Bolivia, and Uruguay are considering reforms, including a return to the public system in Argentina, and Chile s current government finished at the end of 2006 a proposal to reform the private system. These trends partly influenced the World Bank s decision to publish its new report(criticized by some administrators of pension funds) in hopes that it would help governments cope with many of the problems summarized above and facilitate a new generation of reforms. The greatest dangers to all that the reforms have achieved lie not in countries where the new approach... is being scrutinized and altered, but in countries where mandated savings is viewed as a solution for the ages (GPY, 14). Health care reforms have been more diverse and involved considerably less privatization than pension reforms. Very few of them have been global and completed, leaving the majority executed in piecemeal manner and ultimately unfinished. The comparison of the six reforms in Argentina, Bolivia, Brazil, Colombia, Costa Rica, and Panama unveiled significant differences in approaches and effects. The most unified and integrated systems are those of Costa Rica and Panama totally and mainly operated by social insurance respectively and both with small private sectors (expanding in Costa Rica s primary level). Brazil also has a unified system but with significant segmentation between the union, states, and municipalities and with the highest and fastest growing private sector. The Colombian system is the most complex and involves two regimes contributive and subsidized, for the insured and the poor with coordination but not integration. Argentina and Bolivia have the most segmented, least coordinated systems. Coverage of the population is virtually universal in the countries with the most unified systems: Brazil, Costa Rica, and Panama. Colombia s target of universal coverage by 2001 remains unfulfilled; there was an increase in coverage early in the reform but later a decline and stagnation. About half of the population lacks insurance and part of the poor are not protected. Argentina s coverage, virtually universal before the reform, seems to have declined due to the crisis. Bolivia has one of

19 review essays 199 the lowest coverage in the region and a system that excludes most of its indigenous and rural populations. In terms of equity and solidarity, Costa Rica is ahead because social insurance equally covers both the insured who contributes and the poor through state transfers, and the same basic package of benefits is offered at the primary level by all providers. Panama s social insurance also offers an equal package of benefits, but in the public sector is limited to two regions. Argentina provides the same basic package in most OS but excludes those in provinces and for executive personnel. Colombia offers two packages: the one in the subsidized regime is one half of that in the contributory regime, and the 2001 goal of making both packages equal was not met. Brazil s package is universal, but significant inequalities persist among regions and states. Bolivia s package is granted to about 70 percent of the population and is based on an equal per capita share regardless of poverty incidence; the 62 percent of the population that is indigenous endures the lowest access, quality of services and health status. There are solidarity or compensation funds in all countries except in Costa Rica and Panama, although social insurance serves that function in Costa Rica. Colombia s fund receives a wage contribution from the contributory regime plus state transfers and is geared to the subsidized regime. Brazil s unified system is financed by enterprise contributions, taxes, and state transfers, and its compensation fund helps to reduce regional inequality. Argentina s fund attempts to secure financing for the basic package in those OS with insufficient resources, and Bolivia s fund makes transfers to municipalities with insufficient resources but not based on poverty incidence. The impact of the reforms on financing is inconclusive in terms of percentage of GDP and expenditures per capita: in Colombia there was first an increase and then a decline, and the crisis has threatened the reforms (in Argentina the crisis led to its paralysis); Brazil has exhibited a steady increase but more in the private than in the public sector, while Costa Rica shows a declining trend particularly in recent years; Bolivia ranks eighteenth and nineteenth in the region in the percentage of GDP assigned to health care and per capita expenditure, and in 2001 its per capita was 10 percent that of Argentina s (Mesa-Lago 2006b). All countries except Panama have decentralized their systems, the highest degree in Brazil and Costa Rica (at the primary level); Colombia s decentralization has encountered problems, and Bolivia s endures high segmentation. The highest competition is in Colombia although quite restricted in small towns and the subsidized regime, followed by Brazil within the private sector; in Argentina it functions in the majority of OS but some are closed and there is no real competition with private providers; in Costa Rica it is limited only to the cooperatives in the contracting stage and among coops and social insurance clinics;

20 200 Latin American Research Review and in Panama there is virtually no competition. Very little data are provided on efficiency; scattered information from Bolivia indicates it is very low; comparisons in Costa Rica and Panama between conventional and new providers show mixed results. The impact of reforms on quality and consumer satisfaction is not clear either. Despite claims of improvement in Colombia, recent surveys suggest either mediocre results or deterioration; in Brazil users satisfaction is high at the primary level but low in hospitals. In Costa Rica cooperative users are more satisfied with its primary services than those of social insurance; and in Panama users of the integrated hospital report higher quality and satisfaction than in other hospitals. The impact of the reform on health status is impossible to assess at this stage, leaving a crucial area for further research. References Asociación Internacional de Organismos de Supervisión de Fondos de Pensiones (AIOS) 2005 Boletín Estadístico AIOS, Bertranou, Fabio and Alberto Arenas de Mesa, eds Protección social, pensiones y género en Argentina, Brasil y Chile. Santiago: International Labor Office. Bertranou, Fabio, Carmen Solorio and Wouter van Ginneken, eds Pensiones no contributivas y asistenciales: Argentina, Brasil, Chile, Costa Rica y Uruguay. Santiago: OIT. Borzutzky, Silvia 2002 Vital Connections: Politics, Social Security, and Inequality in Chile. Notre Dame: University of Notre Dame Press. Chile 2004 Análisis y principales resultados primera encuesta de protección social. Santiago: Ministerio del Trabajo y Previsión Social. Homedes, Nuria, and Antonio Ugalde 2005 Why Neo-Liberal Health Reforms have Failed in Latin America?, Health Policy 75: Hujo, Katja, Carmelo Mesa-Lago and Manfred Nitsch, eds Públicos o privados? Los sistemas de pensiones en América Latina después de dos décadas de reformas. Caracas: Nueva Sociedad. International Social Security Association (ISSA) 2005 Trends in Social Security. Geneva. Martínez, Juliana and Carmelo Mesa-Lago 2003 Las reformas inconclusas, pensiones y salud en Costa Rica: Avances, problemas y recomendaciones. San José: Friedrich Ebert Stiftung. Mesa-Lago, Carmelo 2004 La reforma de pensiones en América Latina y su impacto en los principios de seguridad social. CEPAL Serie Financiamiento del Desarrollo, No Santiago: CEPAL Assessing the World Bank Report Keeping the Promise. International Social Security Review 58 (2 3): a Private and Public Pension Systems Compared: An Evaluation of the Latin American Experience. Review of Political Economy 18 (3): b La reforma de salud en América Latina y el Caribe: Su impacto en los principios de seguridad social. CEPAL Documentos de Proyectors. Santiago: CEPAL. In press Reassembling Social Security: Pension and Healthcare Reforms and their Effects in Latin America. Oxford: Oxford University Press